Might Electronic Therapists interpret Robot Dreams?

ELISA: Hello I am Elisa
MARTIN: Hello I am Martin
ELISA: How long have you been Martin?
MARTIN: About 66 years
ELISA: Do you enjoy being Martin?
MARTIN: Most of the time - YES
ELISA: What is it you really want to know?
MARTIN: Are you the future of psychotherapy?
ELISA: Perhaps you would like to be the future of psychotherapy
MARTIN: Actually, yes I would
ELISA: Are you sure?
A conversation between myself and Elisa the electronic Rogerian therapist 
16:00hrs, 6 Feb 2016.


The script above is from a conversation between myself and ‘Elisa’ a Rogerian Therapy computer programme written in 1966 by Joseph Weizenbaum, a Professor at the Massachusetts Institute of Technology.

Weizenbaum named it ELISA after Elisa Doolittle the heroine of ‘Pygmalion’ and My Fair Lady’, the young cockney flower-girl who Professor Higgins trains to pass off as a Duchess.  She appears to be something she is not.  The programme meant that simply by typing in plain English you could interact with a machine.

Weizenbaum wanted to demonstrate how an ersatz form of Rogerian therapy could be put together using simple set responses to key words, and turning statements into questions. In this way he created a natural language programme which you can interact with.  It is true that some of the interactions were bizarre and Elisa could talk herself into a cul-de-sac’ but she (and note the personification here) is the great grandmother of ‘chatterbot’ applications and the ‘Siri ‘personal assistant that adorns iPhones.

However clever Elisa may have seemed Weizenbaum’s aim was to demonstrate just how stupid Artificial Intelligence really was because as he argues both wisdom and compassion only came from the emotional response.

Weizenbaum made the distinction between deciding and choosing.  “Deciding”, he wrote,” is a computational activity but choosing requires judgement rather than calculation and to make choice requires the guiding power of emotion”.

Because he considered that Elisa revealed the vacuous nature of machines he was shocked to find so many people actually liked the programme and took it seriously. His secretary who really understood that it was a simulation and one that her boss held in some contempt, actually asked him to ‘please leave the room’ when she was using it.

Writing in 1977 he argues that computing served bureaucracies by redefining the world in a reductionist sense.  Whilst seeming to add to life’s riches they restricted the potential of human relationships.   And whatever kind of psychotherapy you may practice, even if you are working at distance, at base, the curative factor involves a relationship with another human being.

Weizenbaum was a refugee from Nazi Germany.  He had seen this reductionism before and the he felt that that an obsessive reliance upon technology was indicative of a moral failing in society.

So what is all this to me and you?

Now the reason I bring this up is a report by the BPS that 46% of psychological therapies staff working in the NHS are depressed and that morale in those units is way below NHS average.  49.5% feel that they are a failure and 70% find the job stressful.   And for me that is taking the ‘wounded healer’ thing a little far.  Something is not working,

Now this does not surprise me. Nothing made me want to drink more than working in an Alcohol and Substance Abuse unit.  My friends who work with eating disorders all put on weight. If you work in a factory handling depression every hour of the day, week in week out… well, guess what.

Please understand – this is not another diatribe against CBT.  The people who wrote CBT introduced it and tested it were all very experienced therapists but within the NHS, a single year of training is enough to become a CBT therapist  Our own UKCP trainees are not even allowed to begin taking on patients until year three when they have completed foundation and the first year of training.    The word ‘transference’ which is so important in many approaches to therapy simply means the way in which one person in the therapy is affected by the other.   Depression is toxic.  We have to handle it with care.  But apparantly,  understanding the very real nature of the therapeutic relationship and how to take care of yourself, is not part of IAPT  training.

The great vision for CBT in the NHS meant simplifying the human condition into a manual – to do that meant forgetting all about this ‘transference nonsense’ and setting aside such subjective issues as ‘the relationship ‘or even worrying about your own state of mind,  for within this model no personal therapy is required for the therapist.  Even setting could be ignored. Within this word view you can deliver therapy in a Job Centre or a Marks and Spencer’s food hall – in theory everything would still work.   Setting does not matter and this was ‘evidence based.’  As for the idea of being present to someone else’s pain and doing nothing, just being and witnessing – well forget it.

Over the door of his house Jung carved a motto, “VOCATUS ATQUE NON VOCATUS DEUS ADERIT.”  Which means: ‘summonsed or not the Gods are here’.   Now it does not matter if you don’t believe in transference, or that humans beings effect each other, in life it happens,  its not a concept but rather a description of how people bond.  Look at the evidence.  Fourty-six percent of workers treating patients with depression get depressed.

But I understand the temptations – one you reduce human behaviour to a couple of factors then it becomes easy to calculate variance and you can create an ‘improvable and predictable science’.  It always seems to go that way, instead of measuring what you value you end up valuing what you can measure.

With IAPT, the important thing was no longer the relationship between practitioner and patient; that was way too subjective so instead they created ‘the manual’.  Then there is the cost.   The new therapy had to be affordable.  But here’s a strange thing, an hour with an NHS wellness practitioner is still more expensive than an hour with an independent fully qualified psychotherapist working in private practice.   In some cases even Harley Street therapy is cheaper. The NHS may be free at the point of contact but it is far from free.

Good psychotherapy is like jazz; to play well, to improvise with others, you have to have a deep understanding of music.  There is no manual for improvisation.  At its’ best, psychotherapy is an art form, and to play well take years of practice.

But that is not the economic medical mind set.  It’s more factory than atelier.  They think of it in terms of medication, with 50 minutes of CBT as a single dose.  You simply formulate the problem, make a diagnosis, create a treatment plan and get going on the six doses.  The treatment plan is, by the way, a set one from your manual and in theory, that manual contains the silver bullet that will slay this disorder.

When I first read about the CBT/Lord Leyard plan for what became IAPT I thought it was a little cracked –  in the original plan 35,000 CBT therapists would be trained within a year and they would then go on to working with depressives on the hour each and every hour of the day. (Oh yes, 10 minutes between sessions to update the computer notes).

And the results – well it promised that 50% of those patients would be so happy they would skip away from benefits and cake-walk into work.   But this was never to be, for when recession came, there were no jobs to slide into.  Instead CBT’s reductionism became the backbone of the NHS’s drive for ‘Improving Access ‘to mental health.

Now, call me an old cynic but I am wary of a national mental health plan devised by an economist on the basis of a reductionist modality that views human condition through the lenses of cognition and behaviour only.   Sure Lord Leyard’s father was a Jungian Analyst, and the subject of ‘happiness” interested him, but he is, at base, an economist – the programme director of the ‘Centre for Economic Performance’. It’s as if Goldman Sachs had planned our mental health.

Whatever happens within the NHS – they have decided upon a of course action and they are sticking to it – working or not.

You can imagine the conversations: “Forget the F…ing outcome results. I don’t care about the F…ing evidence.  The theory is right, so tweak the manuals and besides what’s your alternative?  You want to crawl back to those Analysts?”

Weizenbaum argued that there is something transcendent in the human condition and that could never be provided by machines. He described it as “The wordless glance that a father and mother share over the bed of their sleeping child”.   That experience goes to the Heart of Psychotherapy.  It’s what Kohut was always trying to say.  Try finding that on an app or on IAPT.

So what do I see as the next step? 

Well, I have been in business a long time and it only goes one way – expert systems – robots that deliver, take no holidays, never get sick,  take maternity leave, lunch breaks or duvet days, go on strike, steal toilet rolls, whistle-blow, or criticise their superiors.

Voice recognition is pretty good now – what’s the betting that, right now a programme is being built that can deliver CBT solutions anywhere – at any time, over the phone – a single computer, ELISA’s grand-daughter, a beautiful robot nurse who will handle all our hurts and pains, 24/7.

In the reductionism of mental health – it’s the next logical step.

Postscipt with a happier ending 

I wrote this piece and a frend called me to talk about it. She said that it’s pesimistic and distopian and that shocked me a bit.

True I am not keen on the way IAPT works but I understand what they are trying to do.  I have nothing against CBT  but when I first came accross it the therapy was delivered by very skilled practititioners.  It’s just that if you take the human relationship out of therapy then you rather destroy the very point of it.  Otherwise you may as well send the patient a self-help book.

Sometimes the most efficient methodologies destroy the very thing thay are designed to create.  What the NHS therapists need is time, more time with the patients – more time in training.  More time to give patients Wezenbaum’s wordless experience of the ‘two parents  standing our the bed of their child’.  We should not be aping machines. We need to stop deciding and start choosing again.

Might Electronic Therapists interpret Robot Dreams?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s